Like breast cancer in women, prostate cancer is a major public health concern in American men. One of the major roadblocks to improving the outcome of this disease is a current lack of understanding of how to accurately determine the risk of disease progression. Without this information, it is difficult to make an informed decisions about what type of treatment is most appropriate, or whether any treatment is needed at all. Although there are many studies in the prostate cancer literature evaluating the diagnostic accuracy of imaging for pretreatment staging using standard pathologic endpoints, there is little data on the ability of imaging to predict tumor aggressiveness or patient outcomes. The broad, long-term objective of this proposed research is to improve the treatment of patients with prostate cancer by the judicious use of diagnostic imaging tests to accurately stage the disease and to help determine the risk of disease progression. Our proposed research investigates a possible new tumor risk factor, which may be used in conjunction with existing risk factors to provide a more accurate assessment of tumor aggressiveness than is currently possible. We wish to test the hypothesis that the morphologic (MRI) and metabolic (MRSI) information provided by MR imaging allows more accurate determination of tumor aggressiveness and prediction of patient outcome than the use of clinical risk factors alone. The specific aims of our study are to: 1. Determine whether the severity of abnormality in metabolism in areas of prostate cancer identified by MRSI represents a significant new independent measure of tumor aggressiveness. 2. Compare the accuracy of MRSI, MRI, and TRUS in determining the local extent of tumor in patients who will undergo radical prostatectomy. 3. Determine the best way to combine the diagnostic information obtained from MRSI, MRI, TRUS, and clinical risk factors to provide more accurate risk assessment than the use of any diagnostic test alone.